AUSTIN, Texas (AP) — A sweeping nonpartisan study released Thursday suggests Texas can be a model for improving juvenile justice systems nationwide, concluding that the state's dramatic shift away from sending youths to detention facilities has coincided with a sharp drop in crime committed by young people.

Compiled by the Justice Center at the Washington-based Council of State Governments, the report found that the number of juveniles held in state centers fell 66 percent between 2007 and 2012, from around 4,305 to about 1,500. Youth crime rates over the same period declined by a third.

Michael Thompson, director of the Council of State Governments Justice Center, said much of the country has moved to reduce the number of juveniles jailed in state facilities and youth crime rates have also declined in many areas— but Texas' saw both fall sharply at the same time.

Lawmakers spent years redesigning the system after pervasive 2007 reports of physical and sexual abuse at youth lockups. As the number of young people confined to state facilities began to fall, many were assigned to community-based programs run by counties and situated closer to their homes.

"Out of sight, out of mind was no longer the standard," said state Sen. John Whitmire, a veteran Houston Democrat who chairs the chamber's Criminal Justice Committee. Of young offenders, he said, "getting them back in their community has been very successful."

The study also found that cutting the number of kids in state-run facilities saved Texas $150 million — much of which was reinvested in community-based programs.

Tracking 1.3 million individual case files, the report noted that Texas continues to have the nation's second-largest juvenile justice system.

It also determined that while re-arrest rates remain too high for all kids who enter the system, those assigned to community programs were 20 percent less likely to be arrested again than those sent to state detention facilities. They were also three times less likely to commit serious offenses.

Less rosy news, however, was the report's finding of relatively little difference in re-arrest rates for juveniles based on the type of community-based program they were exposed to. That meant young people who got treatment were about as likely to be re-arrested as those who simply were placed in surveillance programs.